The TV advertisements tell you it is 98% accurate. The salesmen tell you that you can talk as fast as you normally talk, and in a normal tone of voice. They tell you that with 10 minutes of teaching it your voice, you can be up and running at full speed. Anybody can tell that it has to be much, much cheaper than using a transcriptionist ... or is it?

Seems like every change you would like to make involves a series of tradeoffs. You'd like that sporty new car, but it costs way too much. You'd like to take a two week vacation, but the thought of coming back to all that work that no one did while you were gone makes you think twice about it.

Switching to voice recognition sounds like a no brainer. After all, it is a lot cheaper, and you have your work right away. There's no messing with tapes and no mistakes ... your notes are exactly like you want them. All it takes is for doctor to sit down for a few minutes and voila, it's done - no muss and no fuss, right?

Well, in our experience, that's what the software people would like you to believe, but in reality, it isn't quite so. We have used voice recognition software in our office. We used it out of necessity, to enable us to continue working after a surgery. The software was easy to set up, and relatively easy to train to our voices, but the ease of use ended right there for the most part. They say 95-98% accuracy - the best we could do is perhaps 90%. But, 90% still sounds pretty good, doesn't it? Well, think about that for a moment, that's one error for every ten words you speak. That's a lot of correcting, and correcting words takes time. Of course, you can simply leave the errors and fix them later by hand, or let one of your staff proof and correct it, but that takes time too. But, there are other problems with errors. The software learns from your mistakes, and if you don't correct the errors with the software, it quickly learns to make the same errors over and over.

We had the most difficulty with words that sound the same but have different meanings and spellings. For instance, "to", "too" and "two." The software can't tell the difference between the three words from the context of your speech, so, you generally get the wrong version of the word "to" every time you use the word. You want "too" and you get "to". You want "to" and you get "two". It can be very aggravating.

It's simply impossible to eat a quick sandwich while using voice recognition. And, if the work area is noisy, forget about using it too. Then, there is the big issue of being tied down to a computer for some length of time each day. Aren't there better things to do with your time?

Oh, the last thing you want, is to have a cold and a stuffy nose that changes the tone of your voice. You'll find that the software doesn't remember your voice at all!

We have found that the repetitive phrases a doctor uses tend to be spoken very quickly on a tape and the transcriptionist figures it out. You can't slur a word or a phrase with voice recognition software. You have to speak clearly, purposefully and succinctly. Whatever time is saved whipping through the dictation when using a tape recorder is lost with voice recognition.

It quickly becomes apparent that it takes much longer to do a note using voice recognition than it does to use a tape recorder. It also takes longer to use voice recognition to type a note than it does to actually type it on the keyboard. Using a tape recorder also means doctor's grammar doesn't have to be perfect, he or she can speed right through repetitive words and phrases, eat lunch, dictate in the car, the cafeteria, or at the kitchen table at home. With a tape recorder doctor doesn't have to back up and make corrections and doesn't have to concern himself with using the right form of "to", "too" or "two." And, doctor doesn't have to learn how to use a computer, a wordprocessor, or voice recognition software.

There is no substitute for having a real live, thinking person listening to your dictation who will make corrections, when inadvertently you switch from he to she, or right to left in the middle of a note, when you use the wrong tense, dictate a run-on sentence, or dictate Xanax when you mean Zantac.

In conclusion, while it may sound good on paper, and the numbers certainly look great, there is no substitution for a tape recorder and a transcriptionist to provide the freedom to dictate when and where doctor wants, without the hassle of dealing with a computer after a long day of seeing countless patients and having the satisfaction of knowing that when the tape is transcribed, it will be accurate, polished and complete. Voice recognition takes time away from seeing patients, or time away from the family or from being somewhere else, doing something much more important. So where is the cost savings, really?

Like a nice comfortable chair, or a good bottle of wine, some things in life are simply worth the price. Let doctor dictate, and the transcriptionist transcribe.